Dieting, an attempt to restrict the quantity and/or type of food ingested, is impressively common in the US, especially among women. It is sometimes beneficial, but often ineffective or frustrating, occasionally distressing or even dangerous. Among dieters, 1/3 become pathological dieters and 1/4 of these individuals develop eating disorders. The pursuit of a low-calorie/low-fat diet is a hallmark of a range of eating disorders. central challenge is to understand the cognitive and neurobiological mechanisms underlying the persistent choice of a low-calorie diet and how this behavior transitions across the spectrum of dietary restriction, from potentially health-promoting to potentially deadly. A critical hurdle in addressing this challenge is a long-standing gap between research on the neural underpinnings of choice in healthy individuals and research on caloric restriction in eating disorders. In particular, there have been substantial advances in understanding the mechanisms underlying choices in the healthy brain. But studies of eating disorders have generally focused on characterizing brain responses to passive viewing of food - these studies have not specifically linked neural responses to the restrictive eating behavior itself. To overcome this hurdle, the proposed research program seeks to leverage recent advances in human cognitive neuroscience to understand the neurobiological mechanisms underlying food choice across a full spectrum of dietary restriction. Our specific aim is to examine the neural mechanisms guiding active food choice in individuals whose behavior ranges from healthy to pathological dietary restriction. We will use fMRI combined with a Food Choice Task to measure brain activity directly related to choice behavior in a large sample of individuals, across a spectrum of dietary restriction ranging from healthy controls to healthy dieters, to subthreshold AN and to individuals with full syndrome AN. We will also examine decision-making more broadly, using an experimental paradigm that distinguishes habit-based learning systems. Our study will provide the first data demonstrating how food choice mechanisms function in persistent maladaptive dietary restriction, including among individuals with sub-threshold and full syndrome AN. Our central prediction is that as dietary restriction increases in persistence and severity, food choic preferences are mediated by enhanced cortical self-control mechanisms, and increased dorsal striatal habit mechanisms. The proposed linking of cognitive neuroscience and eating disorders will be a major advance toward elucidating the neurobiology of pathological dietary restriction - a salient feature of eating disorders, and could serve as a model for the study of persistent maladaptive behavior in other serious mental illness.